Feldman et al v. Arizona Secretary of State's Office et alMEMORANDUM of Points and Authorities in Support of 72 MOTION for Preliminary Injunction on Polling Place Allocation and Provisional Ballot ClaimsD. Ariz.June 10, 20161 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Daniel C. Barr (# 010149) Sarah R. Gonski (# 032567) PERKINS COIE LLP 2901 North Central Avenue, Suite 2000 Phoenix, Arizona 85012-2788 Telephone: (602) 351-8000 Facsimile: (602) 648-7000 DBarr@perkinscoie.com SGonski@perkinscoie.com Attorneys for Plaintiffs Roopali H. Desai (# 024295) Andrew S. Gordon (# 003660) D. Andrew Gaona (# 028414) COPPERSMITH BROCKELMAN PLC 2800 North Central Avenue, Suite 1200 Phoenix, Arizona 85004 Telephone: (602) 381-5478 RDesai@cblawyers.com AGordon@cblawyers.com AGaona@cblawyers.com Attorneys for Intervenor-Plaintiff Bernie 2016, Inc. [Additional Counsel Listed on Signature Page] UNITED STATES DISTRICT COURT DISTRICT OF ARIZONA Leslie Feldman, et al., Plaintiffs, v. Arizona Secretary of State’s Office, et al., Defendants. No. CV-16-01065-PHX-DLR MEMORANDUM OF POINTS AND AUTHORITIES IN SUPPORT OF PLAINTIFFS’ JOINT MOTION FOR PRELIMINARY INJUNCTION ON POLLING PLACE ALLOCATION AND PROVISIONAL BALLOT CLAIMS (ORAL ARGUMENT SCHEDULED FOR AUGUST 12, 2016) Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 1 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 TABLE OF CONTENTS Page -i- I. STATEMENT OF FACTS ....................................................................................... 1 A. Arizona’s History of Discrimination and Coverage Under the VRA ........... 1 B. The County’s Allocation of Voting Locations .............................................. 2 C. The Practice of Not Counting OOP Provisional Ballots ............................... 6 II. ARGUMENT ......................................................................................................... 10 A. Plaintiffs Are Likely To Succeed On The Merits ....................................... 11 1. The Challenged Practices Violate Section 2 of the VRA ................ 11 a. Legal Standard ...................................................................... 11 b. County’s Arbitrary Allocation Method Disproportionately Burdens Minority Voters ....................... 13 c. Arizona’s Practice of Not Counting OOP Ballots Disproportionately Burdens Minority Voters ....................... 16 d. Eight of the Nine Senate Factors Are Applicable ................. 18 e. Cumulative Impact ................................................................ 23 2. The Challenged Practices Violate the 14th Amendment ................. 24 a. Allocation of Polling Locations ............................................ 25 b. Out-of-precinct Provisional Ballots ...................................... 26 B. Plaintiffs Will Suffer Irreparable Harm Absent An Injunction ................... 28 C. The Balance Of The Equities Tips Sharply In Plaintiffs’ Favor ................. 28 D. The Public Interest Favors The Issuance Of A Preliminary Injunction ...... 30 III. CONCLUSION ...................................................................................................... 30 Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 2 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -ii- TABLE OF AUTHORITIES Page(s) CASES Anderson v. Celebrezze, 460 U.S. 780 (1983) ...................................................................................................... 24 Ariz. Dream Act Coal. v. Brewer, 818 F.3d 901 (9th Cir. 2016) ........................................................................................ 28 Ariz. Minority Coal. for Fair Redistricting v. Ariz. Indep. Redistricting Comm’n, 211 Ariz. 337 (Ct. App. 2005) ...................................................................................... 27 Beerheide v. Zavaras, 997 F. Supp. 1405 (D. Colo. 1998) ............................................................................... 29 Black v. McGuffage, 209 F. Supp. 2d 889 (N.D. Ill. 2002) ............................................................................ 27 Burdick v. Takushi, 504 U.S. 428 (1992) ................................................................................................ 24, 25 Bush v. Gore, 531 U.S. 98 (2000) ........................................................................................................ 27 Chisom v. Roemer, 501 U.S. 380 (1991) ...................................................................................................... 11 Council of Alt. Political Parties v. Hooks, 121 F.3d 876 (3d Cir. 1997) .......................................................................................... 28 Crawford v. Marion Cty. Election Bd., 553 U.S. 181 (2008) ...................................................................................................... 25 Farrakhan v. Wash., 338 F.3d 1009 (9th Cir. 2003) ................................................................................ 11, 13 Gonzalez v. Ariz., 624 F.3d 1162 (9th Cir. 2010) ...................................................................................... 12 Gonzalez v. Arizona, 677 F.3d 383 (9th Cir. 2012), aff’d sub nom. Ariz. v. Inter Tribal Council of Ariz., Inc., 133 S. Ct. 2247 (2013) ..................................................................... passim Holder v. Hall, 512 U.S. 874 (1994) (Thomas, J. concurring) .............................................................. 12 Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 3 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -iii- TABLE OF AUTHORITIES (Continued) Page(s) Johnson v. Halifax Cty., 594 F. Supp 161 (E.D.N.C. 1984) ................................................................................. 29 Lamb-Weston, Inc. v. McCain Foods, Ltd., 941 F.2d 970 (9th Cir. 1991) ........................................................................................ 10 League of Women Voters of N.C. v. N. Carolina, 769 F.3d 224 (4th Cir. 2014) ............................................................................ 13, 28, 30 M.R. v. Dreyfus, 697 F.3d 706 (9th Cir. 2011) ........................................................................................ 29 McMaster v. United States, 731 F.3d 881 (9th Cir. 2013) ........................................................................................ 23 Melendres v. Arpaio, 695 F.3d 990 (9th Cir. 2012) ............................................................................ 10, 28, 30 Melendres v. Arpaio, 784 F.3d 1254 (9th Cir. 2015) ...................................................................................... 10 Ne. Ohio Coal. for Homeless v. Husted, 696 F.3d 580 (6th Cir. 2012) ........................................................................................ 25 Obama for Am. v. Husted, 697 F.3d 423 (6th Cir. 2012) (OFA) ....................................................................... 28, 30 Oregon v. Mitchell, 400 U.S. 112 (1970) .......................................................................................... 18, 19, 20 Pub. Integrity All., Inc. v. Tucson, 805 F.3d 876 (9th Cir. 2015), reh’g en banc granted, 2016 WL 1696226 (9th Cir. Apr. 27, 2016)................................................................................................. 28 Republican Party of Ark. v. Faulkner Cty., 49 F.3d 1289 (8th Cir. 1995) ........................................................................................ 25 SEC v. McCarthy, 322 F.3d 650 (9th Cir. 2003) ........................................................................................ 23 Shelby County v. Holder, 133 S. Ct. 2612 (2013) ................................................................................................ 1, 2 Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 4 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -iv- TABLE OF AUTHORITIES (Continued) Page(s) Taylor v. La., 419 U.S. 522 (1975) ...................................................................................................... 29 Thornburg v. Gingles, 478 U.S. 30 (1986) .................................................................................................. 12, 13 Williams v. Salerno, 792 F.2d 323 (2d Cir. 1986) .......................................................................................... 28 Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7 (2008) .......................................................................................................... 10 Wood v. Meadows, 207 F.3d 708 (4th Cir. 2000) ........................................................................................ 25 STATUTES 42 U.S.C. § 1973(c) .............................................................................................................. 2 52 U.S.C. § 10301(a) .......................................................................................................... 12 52 U.S.C. § 10301(b) .......................................................................................................... 11 A.R.S. § 16-411 .................................................................................................................... 5 A.R.S. § 16-411(B) ............................................................................................................. 22 A.R.S. § 16-411(B)(4) ...................................................................................................... 2, 7 A.R.S. § 16-583(A) ............................................................................................................. 23 A.R.S. § 16-584(B) ............................................................................................................. 23 A.R.S. § 16-584(C) ............................................................................................................. 23 OTHER AUTHORITIES Lillie Coney, A Call for Election Reform, 7 J. L. & Soc. Challenges 183, 189 (2005) .......................................................................................................................4, 8 Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 5 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -1- For the reasons set forth herein, Plaintiffs request that the Court issue an injunction: (1) requiring the Maricopa County Defendants (the “County”) to make allocation decisions for polling locations for the upcoming General Election in accordance with constitutional requirements and Section 2 of the Voting Rights Act (“VRA”), specifically by enjoining it from approving and implementing an allocation plan for the upcoming November 8, 2016 general election (the “General Election”) that is likely to repeat the types of problems encountered by voters in the recent, disastrous March 22, 2016 presidential preference election (“2016 PPE”); and (2) enjoining Defendants from continuing their practice of not counting provisional ballots cast “out-of-precinct” (“OOP”) in jurisdictions that opt to conduct the General Election under a precinct-based rather than vote center-based model. In support, Plaintiffs submit as follows: I. STATEMENT OF FACTS A. Arizona’s History of Discrimination and Coverage Under the VRA Arizona has a long and deplorable history of racial discrimination against Hispanics, Native Americans, and African Americans, which has permeated every aspect of social, political, and economic life for minorities in Arizona and has been marked, in particular, with restrictions on voting—such as literacy tests and racial gerrymandering— calculated to disenfranchise racial and language minorities. See generally Ex. 1,1 (Expert Report of Dr. David R. Berman (“Berman Rpt.”)); Ex. 2 (Expert Report of Dr. Allan Lichtman (“Lichtman Rpt.”) 23-29). As a direct result, Arizona was one of only nine states to be brought wholly under Section 5 of the VRA as a “covered jurisdiction.” Berman Rpt. 20-21. As such, it was required to obtain “preclearance” of any changes to its elections practices or procedures from either the U.S. Department of Justice (“DOJ”) or a federal court from 1975 until the U.S. Supreme Court issued its opinion in Shelby County v. Holder, 133 S. Ct. 2612, 2631 (2013) striking down the VRA’s formula for identifying covered jurisdictions effectively suspending application of Section 5. While Arizona was 1 Unless otherwise noted, citations to “Ex.” are to exhibits to the Declaration of Elisabeth C. Frost in Support of Plaintiffs’ Motions for Preliminary Injunction. Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 6 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -2- subject to Section 5, DOJ regularly objected to proposed changes to elections practices or procedures in the State, because it determined they had the “purpose” or would have “the effect of denying or abridging the right to vote on account of race or color or [membership in a language minority group].” 42 U.S.C. § 1973(c). In the less than three years that have passed since Shelby County, minority voters have not fared well in Arizona. Both the State and the County have engaged in “consistent activity that has created a culture of voter disenfranchisement” abridging and denying the rights of voters in general and minorities in particular. Ex. 23; see also Berman Rpt. 20- 21. Arizona has also continued to lag behind other states in terms of minority participation in voting. Ex. 3 (Expert Report of Dr. Jonathan Rodden (“Rodden Rpt.”)) 19-21. To this day, the effects of centuries of racial discrimination remain a present reality for Arizona’s Hispanics, Native Americans, and African Americans, who suffer marked disparities as compared to the white population in areas such as employment, wealth, transportation, health, and education. Lichtman Rpt. 38-42; Berman Rpt. 21. B. The County’s Allocation of Voting Locations Arizona law permits counties to choose whether to run each election under a “vote center” model, where voters may cast a ballot at any center in the county, or a precinct- based model, where each voter is assigned to a single precinct and is expected to cast any in-person ballots there. A.R.S. § 16-411(B)(4). Maricopa County has ordinarily held its elections under a precinct model, but the 2016 PPE used a vote center model. Ex. 21. The 2016 PPE represented the first time since 1975 that there was no independent oversight of changes made by the County to its polling allocation scheme and the result was an unmitigated disaster. The County allotted an alarmingly inadequate number of voting centers—just 60 for approximately 1.2 million voters—and the result was severe, as inexcusable burdens were imposed on voters county-wide. Ex. 22 at 5:1-18; Ex. 33; Ex. 23; Ex. 4 (Expert Report of Dr. Muer Yang (“Yang Rpt.”)) 4-5. Voters across the County were forced to stand in line to vote for upwards of three, four, and even five hours. Feldman Decl. ¶¶ 8, 11, 12; Morera Decl. ¶ 7; Magallanes Decl. ¶ 8; K. Gallego Decl. ¶ 6; Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 7 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -3- R. Gallego Decl. ¶ 7, Fernandez Decl. ¶ 2, Quezada Decl. ¶ 7, Clark Decl. ¶¶ 4-5; Healy Decl. ¶ 6, 9, 11; Ex. 23. Thousands remained in line after the polls officially closed at 7:00 p.m. See K. Gallego Decl. ¶ 6; Feldman Decl. ¶ 12; Morera Decl. ¶ 7. Ex. 38. In at least 20 locations, votes were still being cast at 10:00 p.m. Ex. 25; Magallanes Decl. ¶ 8; Healy Decl. ¶ 10. One Phoenix location saw its last voter just two minutes before 1:00 a.m., six hours after the line closed. Ex. 25. Those who were unable to wait (e.g., because of disabilities, transportation issues, or work or family obligations) were disenfranchised altogether. Ovalle Decl. ¶¶ 10, 12; Hymes Decl. ¶¶ 7-9; Ramos Decl. ¶ 5; K. Gallego Decl. ¶6; R. Gallego Decl. ¶ 7, Fernandez Decl. ¶ 2; Quezada Decl. ¶¶ 607; Clark Decl. ¶ 5; Healy Decl. ¶¶ 6, 9, 11; Ex. 23. Voters were also burdened by centers that lacked adequate parking, a problem that itself caused many to be disenfranchised. Feldman Decl. ¶¶ 7, 8; Morera Decl. ¶ 7; Ovalle Decl. ¶¶ 7, 9, 11; Hymes Decl. ¶¶ 8-9; Healy Decl. ¶ 10; Ex. 23. Some locations ran out of ballots, and there was confusion reported over the use of provisional ballots. See, e.g., Feldman Decl. ¶ 11. Hispanic and African-American populations were disproportionately impacted. Rodden Rpt. 4, 61-68. For example, in Senator Quezada’s predominantly Hispanic district, there was only one polling center and many voters—particularly those reliant on public transportation—had to travel long distances to vote. Quezada Decl. ¶¶ 4-6. See also K. Gallego Decl. ¶ 4 (only one vote center to serve predominately Latino area in South Phoenix); R. Gallego Decl. ¶¶ 5-6 (public transportation users in heavily Latino areas faced travel times to polls of up to three hours); Healy Decl. ¶¶ 12-15 (Latino voters lacked access to vote centers and were disproportionately disenfranchised) Ex. 23; Rodden Rpt. 64-68. Some predominantly lower income and Hispanic communities had no polling locations at all. Quezada Decl. ¶¶ 4-6; K. Gallego Decl. ¶ 4; R. Gallego Decl. ¶¶ 4-6; Magallanes Decl. ¶ 7; Healy Decl. ¶¶ 12-13; Ex. 26; Rodden Rpt. 62-63. Maricopa County Recorder Helen Purcell’s (the “Recorder”) explanations how she made the allocation decisions that resulted in these severe and disparate burdens paint a picture of a deeply flawed decision-making process which, if repeated in the General Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 8 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -4- Election, is highly likely to again impose significant, impermissible burdens on Arizonans’ fundamental right to vote. In particular, the Recorder failed entirely to consider how her decisions might impact Arizona’s minority communities and ignored readily available historical data about voter turnout, causing her to vastly underestimate the number of people who would attempt to vote in person on Election Day. See Yang Rpt. 4-8; Ex. 27 at 20:10-21:7.2 She also seriously underestimated the time it would take voters to complete the check-in process, failing to take into account significantly longer time that it takes to check in a voter casting a provisional ballot. Id. at Rpt. 4-5, 9-10. And she assumed, without any reasonable basis, that voters would evenly distribute themselves across voting centers. Id. at 4-5, 10-12. The practical effect was that she failed to adequately allocate voting equipment across voting centers, assuming that four to five check in lines per voting center would be sufficient, when in fact, the turnout at some centers was twice the maximum of what she anticipated per location, and the number of provisional ballots substantially impacted the time that it took to process voters, making five check-in lines woefully inadequate in some locations. Id. at 11-12, 14-15. As a consequence of the County’s disastrous management of the 2016 PPE, many voters have lost confidence in the electoral system. See, e.g., Ramos Decl. ¶ 6; Feldman Decl. ¶ 14, Morera Decl. ¶ 14, Magallanes Decl. ¶ 12; Clark Decl. ¶ 6; Ovalle Decl. ¶ 13, Hymes Decl. ¶ 11, Quezada Decl. ¶ 9. This lack of confidence is particularly heightened in Hispanic communities for which the PPE was just the most recent in a series of actions by the County that have burdened their right to vote. See Berman Rpt. 20. The Recorder has a history of mailing Spanish-speaking (and only Spanish-speaking) voters incorrect information about polling locations, election dates, and even voting propositions. Id.; Ex. 36; Ex. 37; Lillie Coney, A Call for Election Reform, 7 J. L. & Soc. Challenges 183, 189 2 Not only did the Recorder ignore data about turnout in similar elections in her own County, she failed completely to learn from Yuma County’s starkly similar experience only four years prior. See Yang Rpt. 13-14 (discussing Yuma County experience and noting Supervisor Gallardo expressly brought it to the Recorder’s attention during a hearing discussing the allocation plan prior to the PPE). Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 9 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -5- (2005). Such repeated instances of incompetence and neglect have not only created confusion, but have nurtured a culture of distrust of the elections administrators among the County’s Spanish-speaking citizens. Quezada Decl. ¶¶ 9, 27; K. Gallego Decl. ¶ 17; Gillespie Decl. ¶ 16, Healy Decl. ¶ 31; R. Gallego Decl. ¶ 17. Voter fears that they may again encounter the problems they faced in the PPE are entirely rational: virtually all of the errors made by the Recorder could easily be repeated in the General Election. In that fast approaching election, the County will shift back to a precinct-based model. Ex. 24 at 35:19-36:6. The Recorder has stated that she intends to allocate 724 polling places, or one precinct per polling location—which is (except in certain limited circumstances explicitly identified by statute) the bare minimum of polling places in precinct-based elections required by Arizona law. See Ex. 28; A.R.S. § 16-411. Particularly relevant to the errors that she made in planning for the PPE, the Recorder has not explained how she determined that 724 precincts will be adequate in this election; she has not stated whether or how she considered the likely impact on Arizona’s minority voters; she has not explained whether or how she has taken into account how total turnout will vary from precinct to precinct; and she has not indicated whether her calculations incorporate the likelihood of increased registration in the last month before the registration period ends; a number which is likely to vary significantly depending on the precinct. See Yang, Rpt. 16.3 She has similarly not explained how she accounted for the number of voters likely to vote provisionally in each precinct and how that will impact the time it takes to process voters, particularly in precincts with a high number of such voters. Id. Nor has she explained whether she has taken into account any of the following, all of which are relevant to a thoughtful allocation decision: the geographical size of precincts; the availability of public transportation; the proximity of polling places to voters; the characteristics of the facilities that will be used as polling places; the average time that 3 Relatedly, although not an issue in the PPE, but potentially a very serious one in the upcoming General Election is how the implementation of HB2023, which criminalizes the collection of early ballots, could impact turnout and create additional confusion at polling locations. See id. Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 10 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -6- voters will take to complete a ballot in each precinct (likely to vary depending on number of expected provisional voters); the process for curbside voting; the anticipated peak times for voting in each precinct; and the process for providing language assistance to voters who need it. Id. at 16-21.4 Instead, the Recorder’s public stance has been that Arizonans should simply trust that her decisions will adequately serve the needs of the election. But she has not earned that public trust. C. The Practice of Not Counting OOP Provisional Ballots Since 2006, Arizona has rejected over 121,000 provisional ballots, consistently finding itself at or near the top of the list of states that collect and reject the largest number of provisional ballots each election. Ex. 29 at 2; Healy Decl. ¶ 28. In 2012 alone, “[m]ore than one in every five [in person] voters … was asked to cast a provisional ballot, and over 33,000 of these—more than 5 percent of all in-person ballots cast—were rejected. No other state rejected a larger share of its in-person ballots in 2012.” Rodden Rpt. 24-25. See also Ex. 30. Unfailingly, one of the top reasons that ballots are rejected is because they are cast OOP. See, e.g., Rodden Rpt. 26-29. See also Ex. 31 at 5; Ex. 29 at 9. Not all of Arizona’s voters are at risk of being disenfranchised for casting an OOP ballot; the practice of rejecting such ballots only impacts those voters who happen to reside in a county that chooses to run an election under a precinct-based model, rather than a vote center model. In a vote center election, voters may cast their ballots at any 4 Without this information it is nearly impossible to determine whether the County’s decision to allocate the bare minimum of 724 polling places will be adequate or will impose disproportionate burdens on certain voters. For instance, while one polling place per precinct may be appropriate for a precinct with a relatively small geographic area where no voters have to travel far, it may be inadequate for a larger geographic area where some must travel two miles or more to the polling location. See Yang Rpt. 17 (citing authoritative literature finding travel distances to polling locations make “a significant difference to turnout”). One polling station may also be inadequate if a precinct has a large number of voters who typically turn out to vote in person. See id. The number of voters within each precinct also varies widely in the County and the allocation of only one location per precinct is almost certain to impose disparate burdens as a result. Id. at 17-18. If allocation decisions do not take into account the amount of provisional voters likely to appear in each precinct—each of whom will increase the precinct’s average check in time—these burdens are likely to be exacerbated by inadequate distribution of voting equipment. Id. at 18-19. Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 11 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -7- polling place in the county and they will be counted. A.R.S. § 16-411(B)(4). In a precinct- based model, however, only ballots cast by voters in the precinct to which they are assigned will be counted. Ex. 7 at 185-86. For voters unlucky enough to cast a ballot OOP, disenfranchisement is absolute. And the risk that they will do so is high. In this particular area, “Arizona is clearly in a class by itself, with almost 11,000 rejected out-of- precinct provisional ballots in 2012 and 3,500 in 2014. No other state comes close.” Rodden Rpt. 26. While Arizona’s practice of not counting OOP ballots impacts thousands of voters in each election, it disproportionately affects voters in large urban counties, such as Maricopa, which consistently accounts for approximately three quarters of all OOP ballots cast. Rodden Rpt. 28-29; see also Ex. 16 at 6. The practice also disproportionately impacts Hispanic, Native American, and African American voters. In 2012, 70% of in-person voters were white, but “only 56 percent of those casting out-of-precinct ballots were white.” Rodden Rpt. 37. In contrast, only “10 percent of those casting votes at polling places were African American,” but 13% of the ballots cast OOP were by African American voters. Id. A full 26% of OOP ballots were cast by Hispanics, despite the fact that only 15% of in-person voters were Hispanic. Id. Similarly, “[t]he rate at which in- person ballots are [rejected because they were cast OOP] is 80 percent higher for Hispanics, 34 percent higher for African Americans, and 26 percent higher for Native Americans than for whites.” Rodden Rpt. 38-40. Voters with high rates of residential mobility are also disproportionately likely to cast an OOP ballot. Id. at 3-4, 31, 33. Overwhelmingly, these voters are more likely to be minorities. See Lichtman Rpt. 38, 41; Rodden Rpt. 9, 11, 31-32; Quezada Decl. ¶ 25; Healy Decl. ¶ 33; Danley Decl. ¶ 10; Gillespie Decl. ¶¶ 15-16. The primary reasons that voters cast such a substantial amount of OOP ballots in Arizona are twofold. First, voters cast OOP ballots due to their high rates of residential mobility, a factor that is inextricably linked to the State’s long history of discrimination. See infra at 18-21. Second, voters cast OOP ballots due to systemic problems in Arizona’s Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 12 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -8- administration of elections: specifically, voter confusion caused by the large number of changes in polling locations from election to election; the inconsistent election regimes used by and within counties; poor placement of polling locations; and other faulty election administration procedures. Rodden Rpt. 12-15, 26-27, 44-52, 54-58; see also Fernandez Decl. ¶ 19; R. Gallego Decl. ¶ 17; Quezada Decl. ¶ 25; Healy Decl. ¶ 31; Danley Decl. ¶ 10; Gillespie Decl. ¶¶ 15-16; Chapman Decl. ¶ 13; Clark Decl. ¶ 17. In Maricopa County alone, between 2006 and 2008 at least 43% of polling locations changed from one year to the next. Ex. 31 at 4. Likewise, approximately 40 percent of Maricopa County’s active registered voters’ polling locations changed between 2010 and 2012. Rodden Rpt. 56. And “[t]he rate of out-of-precinct voting was 40 percent higher for voters that had experienced a change in polling place.” Id. at 57. See also Ex. 29 at 8; Quezada Decl. ¶ 25; Clark Decl. ¶ 17; Danley Decl. ¶ 10; Healy Decl. ¶ 31. Polling location placement, inconsistent election regimes used by and within counties, and procedural errors are also contributing causes. Rodden Rpt. 12-15, 26-27, 44-52, 54-58; Fernandez Decl. ¶¶ 19-20; Healy Decl. ¶ 30. In 2012, around 25% of individuals casting OOP ballots actually voted at a location closer to their address than the assigned polling place. Rodden Rpt. 4, 53; see also Kelso Decl. ¶ 5; Noonan Decl. ¶ 5. Likewise, “[v]oters assigned to a polling place that is within a mile of two or more other polling places cast out-of-precinct ballots at a rate 13 percent higher than those whose assigned polling place is not in close proximity to other polling places.” Rodden Rpt. 55. In Maricopa County, voters are often confused and need assistance to navigate the differences between the City of Phoenix’s elections (which are vote-center based) and the County’s (usually) precinct-based elections. K. Gallego Decl. ¶ 16; Danley Decl. ¶ 10. See also Rodden Rpt. 4, 55 (residents of Phoenix are more likely to cast OOP ballots). Accounts of elections administrations errors that lead people to cast a ballot OOP abound. See Kelso Decl. ¶ 7, Vasquez Decl. ¶¶ 5-6; Noonan Decl. ¶¶ 6-7; R. Gallego Decl. ¶ 17; Rodden Rpt. 52 (numerous voters disenfranchised because they went to the incorrect table at a consolidated polling location and were not directed to the correct table Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 13 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -9- and/or handed the proper ballot); Lillie Coney, A Call for Election Reform, 7 J. L. & Soc. Challenges 183, 189 (2005); Ex. 21 (spokesman for Recorder’s Office explained that with e-pollbook system OOP votes should not occur at all, yet 2,800 were cast using the e- pollbook system, indicating error). While the high rates of disenfranchisement due to Arizona’s practice of rejecting OOP ballots are alarming in their own right, this problem is likely to be exacerbated in the County in light of the recent PPE debacle, the May 17, 2016 Special Election that was held shortly thereafter, and the city elections that will be held in Phoenix on the same day that the County administers the General Election, where voters will cast their ballots in several contests that are expected to have very large turnout, including the presidential and U.S. Senate races. One of the key factors that has historically contributed to the casting of OOP ballots is the change in the location and number of polling locations. Rodden Rpt. 4, 44-58. As discussed, in the 2016 PPE, 60 voting locations were allotted for all of the County’s voters. In the May 17 Special Election—which was also run under a vote center model where voters could cast their ballots at any polling location in the County—the County allocated 122 polling sites. Rodden Rpt. 15. In the upcoming General Election, the County plans to run a precinct-based election, assigning each voter to only one of 724 polling locations. See Ex. 28. If a voter mistakenly attempts to vote at any of the other 723 polling locations, her ballot will be rejected in its entirety. See id. Further complicating the matter is that the City of Phoenix will also be holding at least one separate city election on November 8, but at entirely different locations and using a vote center model. K. Gallego Decl.¶¶ 18-19; Larios Decl. ¶¶ 5-6; Danley Decl. ¶ 11; Chapman Decl. ¶ 13. While a significant increase in polling locations is plainly necessary to accommodate voters, all of the above is virtually certain to create massive confusion, with minority voters likely to be disparately impacted. See Rodden Rpt. 60-61; See also K. Gallego Decl. ¶¶ 16-19; R. Gallego Decl. ¶ 20; Chapman Decl. ¶¶ 15-16; Clark Decl. ¶ 20-21; Quezada Decl. ¶¶ 28- 29; R. Gallego Decl. ¶ 20; Fernandez Decl. ¶ 19; Healy Decl. ¶ 34; Danley Decl. ¶¶ 11- 12; Gillespie Decl. ¶ 15; Larios Decl. ¶ 6. Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 14 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -10- II. ARGUMENT Plaintiffs are entitled to the requested injunction because (1) they are likely to succeed on the merits of their claims, (2) they are likely to suffer irreparable harm in the absence of relief, (3) the balance of equities tips in their favor, and (4) an injunction is in the public interest. Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008). This Court has “considerable discretion in fashioning suitable relief and defining the terms of a[] [preliminary] injunction,” of which “[a]ppellate review … ‘is correspondingly narrow.’” Lamb-Weston, Inc. v. McCain Foods, Ltd., 941 F.2d 970, 974 (9th Cir. 1991) (quoting Coca-Cola Co. v. Overland, Inc., 692 F.2d 1250, 1256 n.16 (9th Cir. 1982)). Although the injuries suffered by voters in the 2016 PPE cannot be rectified, the Court has the power to protect Plaintiffs and all of the County’s voters in the General Election by enjoining the County from approving and implementing an allocation plan that is likely to repeat the types of problems encountered by voters in the PPE. Plaintiffs are not requesting that the Court override the County’s discretion to make reasonable allocation decisions, only that it require the County to make those decisions in accordance with the VRA and the Constitution, to ensure that voters are not again subject to unjustified burdens on their fundamental right to vote that are the direct result of the County’s failure to take into account clearly relevant information. See, e.g., Melendres v. Arpaio (“Melendres I”), 695 F.3d 990, 995-96, 1000-02 (9th Cir. 2012) (affirming preliminary injunction enjoining defendant Maricopa County Sheriff’s Office from detaining any one based solely on reasonable suspicion or knowledge they are unlawfully present in the country, given plaintiffs’ likelihood of success on constitutional claims and presence of other Winter factors). See also Melendres v. Arpaio, 784 F.3d 1254, 1265 (9th Cir. 2015) (district courts have “broad discretion” in tailoring injunctive relief and are “permitted to order ‘relief that the Constitution would not of its own force initially require if such relief is necessary to remedy a constitutional violation’”). Likewise, the Court has the power to ensure that thousands of Arizona voters are no longer completely disenfranchised as a result of the state’s arbitrary and burdensome Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 15 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -11- practice of not counting OOP ballots by enjoining the wholesale rejection of such ballots, and requiring the State to count all of the elections on an OOP provisional ballot for which the voter would have been eligible to vote had she cast a regular ballot in her assigned precinct. A. Plaintiffs Are Likely To Succeed On The Merits 1. The Challenged Practices Violate Section 2 of the VRA Both the County’s arbitrary allocation decisions and the State’s policy of rejecting OOP ballots wholesale violate Section 2 of the VRA, because each independently and cumulatively results in Arizona’s Hispanic, Native American, and African-American voters having “less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.” 52 U.S.C. § 10301(b). a. Legal Standard. Section 2 is violated “if, based on the totality of the circumstances, it is shown that the political processes … in the State … are not equally open to participation by members of” a racial or language minority group “in that its members have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.” 52 U.S.C. § 10301(b). “Congress enacted the [VRA] for the broad remedial purpose of rid[ding] the country of racial discrimination in voting,” and it must be interpreted so as to “provide[] the broadest possible scope in combating racial discrimination.” Chisom v. Roemer, 501 U.S. 380, 403 (1991) (citations and internal quotation marks omitted). Thus, proving a Section 2 claim “does not require any showing of discriminatory intent, only discriminatory results.” Gonzalez v. Arizona, 677 F.3d 383, 405 (9th Cir. 2012), aff’d sub nom. Ariz. v. Inter Tribal Council of Ariz., Inc., 133 S. Ct. 2247 (2013). “[I]n examining the totality of the circumstances to determine whether a challenged voting practice results in [discrimination] on account of race,” the court “consider[s] how the practice ‘interacts with social and historical conditions to cause an inequality in the opportunities enjoyed by [minority] and white voters to elect their preferred representatives.’” Farrakhan v. Wash., 338 F.3d 1009, 1016 (9th Cir. 2003) Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 16 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -12- (quoting Thornburg v. Gingles, 478 U.S. 30, 47 (1986)). It is this interaction that provides the causal connection necessary to demonstrate a Section 2 violation, i.e., “plaintiff need not show that the challenged voting practice caused the disparate impact by itself.” Gonzalez v. Ariz., 624 F.3d 1162, 1193-94 (9th Cir. 2010) (citation omitted). Likewise, plaintiffs need not show that the practice makes voting impossible for minorities—only that it makes voting disproportionately more burdensome. See Thornburg v. Gingles, 478 U.S. 30, 35-36, 44, 47 (1986); Holder v. Hall, 512 U.S. 874, 922 (1994) (Thomas, J. concurring). Section 2 thus prohibits not only the outright “denial,” but also the “abridgement” of the right to vote. 52 U.S.C. § 10301(a). In determining whether a challenged practice works together with the totality of the circumstances to deny minority voters equal opportunities, courts look to nine non- exclusive factors known as the “Senate Factors:” Senate Factor 1: The history of voting-related discrimination in the jurisdiction; Senate Factor 2: The extent to which voting in the jurisdiction’s elections is racially polarized; Senate Factor 3: The extent to which the jurisdiction has used voting practices or procedures that tend to enhance the opportunity for discrimination against the minority group; Senate Factor 4: The exclusion of members of the minority group from candidate slating processes; Senate Factor 5: The extent to which minority group members bear the effects of past discrimination in areas such as education, employment, and health, which hinder their ability to participate effectively in the political process; Senate Factor 6: The use of even subtle racial appeals in political campaigns; Senate Factor 7: The extent to which members of the minority group have been elected to public office in the jurisdiction; Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 17 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -13- Senate Factor 8: Evidence demonstrating that elected officials are unresponsive to the particularized needs of the members of the minority group; and Senate Factor 9: The extent to which the policy underlying the use of the contested practice is tenuous. Farrakhan, 338 F.3d at 1015-16; see also Gingles, 478 U.S. at 36-37. “‘[T]here is no requirement that any particular number of factors be proved, or that a majority of them point one way or the other.’” Gonzalez, 677 F.3d at 406 (quoting Gingles, 478 U.S. at 45). In giving the VRA its broadest possible scope, courts must “engage in a ‘searching practical evaluation of the “past and present reality,”’” id. (citing Gingles, 478 U.S. at 45), that is “‘intensely fact-based and localized,’” id. (quoting Smith v. Salt River Project Agric. Improvement. & Power Dist., 109 F.3d 586, 591 (9th Cir. 1997)). See also League of Women Voters of N.C. v. N. Carolina (“LOWV”), 769 F.3d 224, 241 (4th Cir. 2014) (“Clearly, an eye toward past practices is part and parcel of the totality of the circumstances.”). The number of voters denied equal electoral opportunity by a law is not relevant. “[W]hat matters … is … simply that ‘any’ minority voter is being denied equal opportunities.” LOWV, 769 F.3d at 244 (quoting 52 U.S.C. § 10301(a)). b. County’s Arbitrary Allocation Method Disproportionately Burdens Minority Voters The County’s arbitrary and unsupportable allocation decisions in the 2016 PPE plainly violated Section 2 and, without the issuance of the requested injunction, there is a strong likelihood that similar mistakes will again be made in the General Election, denying minority voters of an equal opportunity to participate and to elect the representatives of their choice. Although much of the media attention in the aftermath of the 2016 PPE was on the Recorder’s plainly unsupportable decision to allocate only 60 polling locations for all of the County’s voters, equally troubling was her admitted failure to so much as consider the impact that her decisions were likely to have on Arizona’s minority voters, a failure that was reflected in the highly inequitable distribution of those polling locations. See Ex. 27 at Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 18 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -14- 20:10-21, 43:18-44:16; Ex. 23; Ex. 27 at 20:10-21:7; Ex. 32; Ex. 26 at 1-2; Quezada Decl. ¶¶ 4-7; K. Gallego Decl. ¶ 4, R. Gallego Decl. ¶¶ 5-6; Healy Decl. ¶¶ 12-15. This inequitable distribution disproportionately reduced the ability of Hispanics and African Americans to participate in the PPE. Rodden Rpt. 4-5, 61-68. In particular, heavily Hispanic sections of Western Phoenix, Glendale, and Mesa were significantly underserved in comparison to the rest of the County. Rodden Rpt. 62; see also Quezada Decl. ¶ 4; Healy Decl. ¶¶ 12-13; Ex. 26 at 2. On the whole, Hispanics and African Americans had to endure more “costs,” i.e., burdens, to vote in the PPE.5 In addition to the burdens of long wait times, inadequate parking, and poor election administration that all voters who participated (or attempted to participate) had to bear, Hispanic and African American voters endured disproportionate travel burdens when compared to other voters. Rodden Rpt. 61-68. Both in straight line distance and actual travel times, Hispanics and African Americans had to travel farther to reach the 2016 vote centers than they have had to travel to reach their polling locations in prior elections as compared to white voters. Id. “For instance, while [only] 22 percent of non-Hispanic whites experienced an increase of more than 2 miles [in their travel distance for voting], [over] 25 percent of African Americans and 30 percent of Hispanics” did. Id. at 64. An increase in excess of two miles is significant: prevailing political science literature finds that a change of more than two miles is likely to have a substantial negative impact on voter participation. Id. The elimination of several polling places across metro Phoenix also made it substantially more difficult for Phoenix voters who rely on public transportation to vote. 5 It is well established that voting entails both costs and benefits and, when the costs to a voter outweigh the benefits, the voter will abstain from voting. Rodden Rpt. 7-8; Lichtman Rpt. 38. Because minority voters often have fewer resources than white voters, they are generally more susceptible to “costs” placed on their ability to vote. Rodden Rpt. 9; Lichtman Rpt. 38. Some of the most commonly recognized costs associated with voting are: (1) travel costs—i.e., the effort necessary to travel to a polling location; (2) search costs—i.e., the effort necessary to locate or learn where one’s polling location is located; and (3) information costs—i.e., the time and effort necessary to educate oneself about an election, voting laws, or regimes. See Rodden Rpt. 7-11; Lichtman Rpt. 38. Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 19 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -15- Id. at 65-68. Hispanics and African Americans are less likely to have access to a vehicle and more likely to rely on public transportation than whites in Arizona. Lichtman Rpt. 38, 41; Rodden Rpt. 65. And the evidence supports the conclusion that Hispanic voters were forced to travel greater distances, had increased travel time, and were disproportionately more likely to be disenfranchised due to the lack of accessible means of transportation to PPE polling locations. See, e.g., Quezada Decl. ¶¶ 4-7; Healy Decl. ¶¶ 12-13; K. Gallego Decl. ¶¶ 3-4, 6; R. Gallego Decl. ¶¶ 4-7. Hispanic voters also were burdened by greater “search costs” than white voters, because they were more likely to have to search for a new polling location since their “usual” polling locations, i.e., 2014 mid-term polling locations, were less likely to be used in the PPE. Rodden Rpt. 62-63. Hispanic and African-American voters are impacted more heavily than whites by search costs and increased search costs correlate directly with decreased turnout. Id. at 9, 60-61. These costs are particularly acute for Spanish-speaking voters because they must navigate much of the materials designed to educate them on voting in an unfamiliar language. Larios Decl. ¶ 4; Chapman Decl. ¶¶ 4, 13; K. Gallego Del. ¶ 17; Gillespie Decl. ¶ 16; R. Gallego Decl. ¶ 17; Quezada Decl. ¶ 27; Danley Decl. ¶ 5. These additional burdens, when taken together with the extremely long wait times, significantly increased the overall cost of voting for Hispanic and African American voters as compared to whites, who did not incur the same travel or search costs. Rodden Rpt. 60-68; Quezada Decl. ¶ 7; K. Gallego Decl. ¶¶ 3-4; R. Gallego Decl. ¶ 4. As a result, Hispanic and African-American voters faced disproportionate burdens and were significantly more likely to be deterred from voting. There is substantial evidence that participating in the PPE was in fact made extraordinarily difficult or impossible for many Hispanic voters due to a lack of available transportation. See, e.g., Healy Decl. ¶ 13, 33; Quezada Decl. ¶¶ 3-7; K. Gallego Decl. ¶¶ 3-4; Shapiro Decl. ¶ 7. Thus, the cost of traveling a farther distance to vote and waiting in line can be—and in the PPE, in many cases, demonstrably was—a highly effective means of disenfranchisement. See Quezada Decl. ¶¶ 6-7; K. Gallego Decl. ¶¶ 3-4; R. Gallego Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 20 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -16- Decl. ¶¶ 5-7; Healy Decl. ¶ 13. As discussed infra at 18-21, as a direct result of the State’s extended history of discrimination, Arizona’s Hispanic citizens often hold less flexible jobs and are more dependent on hourly wages (rather than salaries), making it substantially more costly (and in some cases prohibitively so) for them to miss work, even for a single shift. See Lichtman Rpt. 38-39; Ramos Decl. ¶¶ 4-6; Quezada Decl. ¶ 3. And there is substantial evidence that many Hispanic voters were disenfranchised in the PPE because inflexible work schedules made it impossible for them to wait in line for hours to vote. See, e.g., Healy Decl. ¶ 14; Ramos Decl. ¶¶ 4-5; Ovalle Decl. ¶ 10. Minorities in Arizona also have much poorer health outcomes than whites, making it disproportionately more burdensome for them to have to endure excessively long lines in order to vote. Lichtman Rpt. 38, 42; Healy Decl. ¶¶ 10-15; Morera Decl. ¶ 1; Magallanes Decl. ¶ 10; Fernandez Decl. ¶ 5; K. Gallego Decl. ¶ 6. Unless the County is required to specifically take into account how its polling allocation decisions are likely to impact Arizona’s minority voters for the General Election, the same communities are highly likely to again face disproportionate burdens which, viewed in light of the totality of the circumstances as demonstrated by the Senate Factors discussed infra at 18-23, run in violation of Section 2 of the VRA. c. Arizona’s Practice of Not Counting OOP Ballots Disproportionately Burdens Minority Voters. Arizona’s practice of not counting OOP ballots violates Section 2, and without the issuance of the requested injunction, will continue to disproportionately disenfranchise minority voters, resulting in minority voters having fewer opportunities to participate and elect representatives of their choice. Minorities are “vastly over-represented among those casting out-of-precinct ballots” and, as a consequence, far more likely to be disenfranchised as a result. Rodden Rpt. 37. In 2012 in Maricopa, while 70% of voters who voted in person were white, “only 56 percent of those casting ballots classified out-of-precinct were.” Id. In contrast, “[w]hile 10 percent of those casting ballots classified out-of-precinct were African Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 21 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -17- American, 13 percent of the ballots invalidated due to” OOP voting were cast by African Americans. Id.; see also id. at 38-40. Similarly, while only 15% of all in-person voters “were Hispanic, 26 percent of the invalid [OOP ballots] were cast by Hispanics.” Id.; see also id. at 38-40; Ex. 34 at 11; Ex. 29 at 13. Further, of all provisional ballots cast in 2012 the rate of OOP voting was 37% higher for Native Americans than for whites. Rodden Rpt. 38-40; see also Ex.33 at 14. These disparities “have been quite persistent over time.” Rodden Rpt. 37-38. The disparate burdens of Arizona’s OOP practice are also linked to the ongoing effects of Arizona’s history of discrimination. OOP voting “is far more pronounced in corridors with more renters, and less common in areas with high rates of home ownership.” Rodden Rpt. 31. This is unsurprising. Minorities, as a result of Arizona’s discriminatory history, are far more likely to rent than own a home. Lichtman Rpt. 38, 41; Rodden Rpt. 31-32; see also Ex. 31 at 4; Danley Decl.¶¶ 5, 10; Healy Decl. ¶ 33; Quezada Decl. ¶ 25. These minority voters, therefore, bear more costs to have their vote counted in a precinct-based voting system that rejects OOP votes—they must reregister with each move (even short, across town moves); continuously reeducate themselves about their new voting location; and be informed that their vote will only count if they vote in their new location. See Lichtman Rpt. 38; Rodden Rpt. 7-12 (discussing, among other things, increased problems for minorities overcoming search costs). Consequently, they are more likely to vote in the incorrect precinct. See discussion supra 6-8. Where a minority voter arrives at the wrong precinct, they are also less likely to be able to remedy the error (assuming they are advised that they must go to their assigned precinct to cast a countable ballot). Minorities have less access to vehicles and are more likely to rely on public transportation or assistance to travel. Lichtman Rpt. 38, 41; Rodden Rpt. 65; see also Shapiro Decl. ¶ 7; Quezada Decl. ¶¶ 3-7; K. Gallego Decl. ¶¶ 3-4; Healy Decl. ¶ 33.Shapiro Decl. ¶ 7; Healy Decl. ¶ 33. Similarly, minorities disproportionately hold less flexible, working-class jobs that can make it more difficult for them to a second location before the polls close. Lichtman Rpt. 38-39; Healy Decl. ¶ 33; Ramos Decl. ¶ 5. All of Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 22 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -18- these facts strongly support a finding that Arizona’s minority voters are disproportionately disenfranchised by the State’s policy of rejecting OOP ballots—decreasing their overall ability to participate in elections and elect the candidate of their choice. d. Eight of the Nine Senate Factors Are Applicable The disproportionate burdens discussed above, when considered in light of the totality of the circumstances, strongly support Plaintiffs’ Section 2 claims. At least eight of the nine Senate Factors—each directly relevant to the Court’s analysis—are present. Lichtman Rpt. 2. They are discussed briefly below: History of Discrimination and Discriminatory Use of Voting Practices (Senate Factors 1 and 3). Arizona has a long, recognized history of discrimination against Hispanics, Native Americans, and African Americans that has extended to every area of social, political, and economic life, including voting. See generally Berman Rpt.; Gonzalez, 677 F.3d. at 406-07. Shortly after becoming a state in 1912, Arizona enacted a literacy test for voter registration designed “to limit ‘the ignorant Mexican vote.’” Berman Rpt. 12. Native Americans were formally excluded from voting in Arizona until 1948, a full 24 years after federal law allowed them to vote. Id. at 15. Even after being granted the right to vote, the literacy test, targeted voter suppression and intimidation tactics, and election administration decisions directed at Native Americans—such as the frequent movement of polling locations—worked to keep them from voting well into the 20th century. Id. at 15-16. In 1970, Congress found that, “[in] Arizona, only two counties out of eight with Spanish surname populations in excess of 15% showed a voter registration equal to the state-wide average” and that Arizona also had a serious deficiency in Native American voter registrations. Id. at 13. See also Oregon v. Mitchell, 400 U.S. 112, 132 (1970). It took Arizona until 1972—two years after Congress enacted a nationwide ban on literary tests and after Supreme Court rejected Arizona’s attempt to fight that ban in court—to repeal its literacy test and restrict its use. Berman Rpt. 12. Discriminatory elections practices aimed at minority groups have continued in recent decades. While Arizona was subject to preclearance, DOJ vetoed at least four Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 23 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -19- redistricting plans and more than 80% of all of its objections to voting practices from the 1980s onward involved voting changes that appeared to threaten the turnout of Hispanics and Native American voters with limited English proficiency. Id. at 21. In just 2000, Arizona banned bilingual education—making it significantly more difficult for Hispanics and Native Americans to gain the literacy skills needed to vote; in 2004, it adopted Proposition 200, which requires the presentment of proof of citizenship when registering to vote and has made it more difficult for both Hispanics and Native Americans to do so; and in 2010 it enacted SB1070, making it a crime for any person who is not a U.S. citizen to be caught without their immigration documents in Arizona, and which resulted in the racial profiling and unreasonable detention of many of Arizona’s Hispanic citizens. Id. at 18-19; Lichtman Rpt. 23-29. In recent years, the County has repeatedly failed to send properly translated education materials to its Spanish speaking residents, resulting in confusion and distrust from Hispanic voters for the County. Berman Rpt. 20; Lichtman Rpt. 50; see also discussion supra at 5. On-Going Effects of Discrimination and Lack of Responsiveness (Senate Factors 5 and 8). From its earliest beginnings, Arizona has limited minority access to education, employment, and public life, and disregarded protections for its most vulnerable citizens. See generally Berman Rpt.; see also Lichtman Rpt. 23-29. The effects of this storied and systemic discrimination persist today, impacting social, economic, and political life of Hispanics, Native Americans and African Americans in the following ways: Poverty: The poverty rate for African Americans and Hispanics in Arizona is nearly two times that of whites. Lichtman Rpt. 39 (23.5% of African Americans and 29.1% of Hispanics and only 11.6% of whites live below the federal poverty line). The poverty rate for Native Americans is over three times higher than that of whites. Id. (39.8% of Native Americans live below the federal poverty line). Employment. Since whites first settled in Arizona, Hispanics, African Americans, and Native Americans have received lower wages for the same work, and have received little to no civil rights protections. Berman Rpt. 6, 9-10; Lichtman Rpt. 39 (discussing the passage of a law in 1996 which made it more difficult for minorities to bring racial discrimination lawsuit in employee termination). Whites have a Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 24 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -20- significantly lower unemployment rate (7.6%) than Hispanics (12.5%), African Americans (13.2%), and Native Americans (19.3%). Id. Education. From 1909 to the 1950s, racial and language minorities were forced to attend segregated schools. Berman Rpt. 7-8. In the post-desegregation era, Arizona has systemically underfunded its minority education programs. Lichtman Rpt. 45- 46. Whites are more likely than Hispanics, Native Americans, and African Americans to graduate from high school. Id. at 40. Whites are nearly three times more likely to have a bachelor’s degree (33.2%) than Hispanics (10.7%) and Native Americans (10%). Id. Consistent with Arizona’s long history of educational discrimination, including its ban on bilingual education programs and underfunding of its English education programs, in a recent survey, over 24.6% of Hispanics and 10.1% of Native Americans rated themselves as speaking English less than “very well” as compared to only 1% of whites. Id. Transportation. Whites are also significantly more likely to own a vehicle than Hispanics, Native Americans, and African Americans. Id. at 38, 41. A recent survey found that only 5.5% of whites in Arizona did not own a vehicle as compared to over 17% of Native Americans, 16% of African Americans, and 8.4% of Hispanics. Id.; see also Quezada Decl. ¶ 3; K. Gallego Decl. ¶ 3. Residential Transiency. While 69.3% of whites own a home in Arizona, only 33.8% of African Americans, 50.8% of Hispanics, and 57.1% of Native Americans do. Lichtman Rpt. 41. Thus, Hispanics, African Americans, and Native Americans are more likely to move than whites and have higher transiency rates. See Rodden Rpt. 7-12; Quezada Decl. ¶¶ 3, 25; Shapiro Decl. ¶ 7; Danley Decl. ¶¶ 3-4, 10. Health. Hispanics, Native Americans, and African Americans also fare far worse than whites on a number of health indicators. Native Americans, African Americans, and Hispanics are significantly more likely to die before the age of 65 than whites. Lichtman Rpt. 42. All have higher infant mortality rates and lower average birth weights than whites. Id. A recent study rated Native Americans, African Americans, and Hispanics worse overall than whites on a measure of 69 health variables. Id. These ongoing effects of discrimination are highly relevant to a Section 2 analysis because “[d]ecades of research has demonstrated that socio-economic standing significantly impacts the ability to fully participate in the political process.” Id. at 38. Minority voters who suffer from health problems may find it more difficult to wait in lines at polling locations and, in some instances, to travel to them altogether. Lichtman Rpt. 27, 38. Voters with limited access to vehicles who rely on public transportation or other Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 25 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -21- assistance may also be impeded from traveling to the polling location in the first instance, or to the correct polling location should they first go to the wrong one. Id.; Rodden Rpt. 61-66; see also Quezada Decl. ¶¶ 3-6, 26; Healy Decl. ¶ 33. English and educational differences may make it more difficult for voters to fully educate themselves on the voting process, including, the location at which they vote; due dates for absentee ballots; or restrictions on where they can vote. Lichtman Rpt. 27, 38; see also Danley Decl. ¶ 5; Quezada Decl. ¶¶ 9, 27; Gillespie Decl. ¶ 16, Chapman Decl. ¶¶ 4, 13; Larios Decl. ¶ 4; K. Gallego Decl. ¶ 17; Healy Decl. ¶ 31; R. Gallego Decl. ¶ 17. Racially Polarized Voting, Racial Appeals, and Minority Electeds (Senate Factors 2, 6, 7). Arizona has a demonstrated history of racially polarized voting. Gonzales, 677 F.3d at 406-07; see also Lichtman Rpt. 29-32; Berman Rpt. 9-10. And, campaigns in Arizona—both historically and recently—have been characterized by subtle and overt racial appeals. When Raul Castro, Arizona’s only Hispanic to ever be elected to a statewide office, ran for governor in the 1970s, his opponents urged support for the white candidate because “he looked like a governor.” Berman Rpt. 17. In that same election, a newspaper published a picture of Fidel Castro with a headline that read “Running for governor of Arizona.” Id. In a 2010 bid for State Superintendent of Public Education, John Huppenthal “ran an advertisement in which the announcer said that Huppenthal was ‘one of us.’ The announcer noted that Huppenthal voted against bilingual education and ‘will stop La Raza.’” Lichtman Rpt. 43. Similarly, when running for governor in 2014, Maricopa County Attorney Andrew Thomas ran an ad describing himself as “the only candidate who has stopped illegal immigration” while “simultaneously show[ing] a Mexican flag with a red strikeout line through it superimposed over the outline of Arizona.” Id. Likewise, at a recent Arizona rally, the presumptive Republican presidential nominee, Donald Trump, made negative statements about Hispanics. Id. at 43-44. All of this has resulted in a marked disparity in the number of minority elected officials. Only one Hispanic and one African American have ever been elected to Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 26 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -22- statewide office in Arizona and the State has never elected a Native American to statewide office. Lichtman Rpt. 44-45. Hispanics, Native Americans, and African Americans are also underrepresented in Arizona’s judiciary and state legislature. Id.; Berman Rpt. 21. Thus, it is clear that the relevant historical and social conditions are such that they interact with the challenged voting restrictions to impose a disproportionate burden on the ability of Hispanics, Native Americans, and African Americans to participate in and elect the candidates of their choice. Tenuous Justifications (Senate Factor 9). The County’s likely explanation for its allocation decisions in the 2016 PPE and the State’s rationale for discarding OOP ballots are both highly tenuous. The County will likely assert that Plaintiffs’ polling allocation claims are unwarranted because the County has already stated that it will have more polling locations in the General Election. But, as explained, many of the same faulty assumptions underlying the Recorder’s allocation decisions for the 2016 PPE could easily be repeated in the General Election with similarly disastrous results for Arizona voters, in particular minorities. See supra at 13-16; see also Yang Rpt. 4, 22. In particular, in the PPE the Recorder (1) ignored critical historical in-person turnout data; (2) underestimated check-in times for voters and insufficiently allocated e-pollbooks for check-in; (3) assumed (with no basis) that voters would distribute evenly across polling locations; and (4) failed to consider the impact of her allocation decision on minority voters. Id. at 4, 21- 22; Ex. 27 at 20:10-21:7. There is no indication that the Recorder has corrected these assumptions. In fact, given that she has only agreed to the minimum numbers of polling locations required under law; has stated that there will only be one polling location per precinct—regardless of the geographic size or number of voters in the precinct; and has given no indication that she will consider the impact of her determination on minority voters, it appears that the Recorder is actually making many of the same faulty assumptions that injured minority voters in the 2016 PPE and that the County’s rationales are, therefore, tenuous. Yang Rpt.16-19; A.R.S. § 16-411(B). Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 27 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -23- As for the State’s practice of not counting OOP ballots, it will likely argue that the law requires it. But this is not supportable. The two key statutory provisions addressing OOP provisional ballots provide only that: an election official “shall” direct a registrant who has moved “to the polling place for the new address,” A.R.S. § 16-583(A), and that a registered voter is permitted to vote a provisional ballot upon “signing an affirmation that states that the elector is a registered voter in that jurisdiction and is eligible to vote in that jurisdiction.” A.R.S. § 16-584(B). The term “jurisdiction” in the latter cannot mean the voter’s precinct, because the term “precinct” is separately used earlier in the very same sentence, indicating that the terms have two different meanings. See e.g., SEC v. McCarthy, 322 F.3d 650, 656 (9th Cir. 2003) (“It is a well-established canon of statutory interpretation that the use of different words or terms within a statute demonstrates that Congress intended to convey a different meaning for those words.”); McMaster v. United States, 731 F.3d 881, 893 (9th Cir. 2013) (different words used within statute presumed to carry distinct meanings). See also A.R.S. § 16-584(C) (similar use of jurisdiction versus precinct). Thus, the State’s policy of rejecting OOP ballots is an improper interpretation of the law, see Ex. 7 at 185, that goes far beyond the relevant statutory provisions. Accordingly, any rationale that state law requires that such ballots be rejected is plainly tenuous. e. Cumulative Impact. Even before the 2016 PPE, elections in Arizona were fraught with systemic problems that lead to the disparate disenfranchisement of minority voters, particularly as a result of Arizona’s practice of not counting OOP provisional ballots. See Rodden Rpt. 4, 22-29; Ex. 16 at 1. Post-PPE, the risk of disenfranchisement has only increased. Many of the same voters who have had problems adjusting to changes in polling locations and navigating voter information in the past, see, e.g., Rodden Rpt. 58-59 (turnout for voters whose polling locations changed from 2012 PPE to 2012 General Election was 30% lower than turnout for voters whose polling locations did not change), will now not only have to locate a new polling location, but they will also have to navigate an entirely different— Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 28 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -24- more restrictive—voting system than the one encountered in the PPE. In the PPE, a voter could vote anywhere in the County, now voters can only vote in their assigned precinct. Clark Decl. ¶ 20; Quezada Decl. ¶¶ 28-29; R. Gallego Decl. ¶ 19; Healy Decl. ¶ 31; Fernandez Decl. ¶ 22; Chapman Decl. ¶ 15; Larios Decl. ¶ 6; Danley Decl. ¶ 11. Navigating these changes will be particularly cumbersome for minority voters who, historically, have more difficulty adjusting to electoral changes, see Litchman Rpt. 38, and who were most impacted by the changes in the PPE. Rodden Rpt. 4, 61-68. This will be further complicated by the fact that Phoenix is running at least some city elections on the same day, at different locations, under a vote center model. Rodden Rpt. 15; K. Gallego Decl. ¶ 19. Finally, the recent passage of HB 2023, discussed in Plaintiffs’ separate motion for a preliminary injunction, contributes to the cumulative effects of both the County’s allocation failures and the State’s OOP policy. Criminalization of ballot collection, relied upon disproportionately by Arizona’s minority voters, see generally Pls.’ Mot. for PI on HB2023 Claims; Gallardo Decl. ¶¶ 5, 18, 22,6, makes it more likely that there will be more people at the polls, creating longer lines and making it more likely that, unless thoughtfully considered, the Recorder’s allocation plans will be insufficient. See Yang Rpt. 6-8. 2. The Challenged Practices Violate the 14th Amendment Plaintiffs are also likely to prevail on their claim that the election practices challenged in this litigation (independently and in concert) impose severe and unjustified burdens on the right to vote in violation of the 14th Amendment. The Supreme Court has developed a balancing test to determine whether facially nondiscriminatory elections laws impose an “undue” burden on voters in violation of the 14th Amendment. Anderson v. Celebrezze, 460 U.S. 780, 789 (1983); Burdick v. Takushi, 504 U.S. 428, 434 (1992). In applying that test, courts “weigh ‘the character and magnitude of the asserted injury to the 6 Declarant Steve Gallardo is named as a Defendant in his official capacity in this litigation. Plaintiffs’ counsel communicated with Gallardo to prepare his declaration with the express permission of his counsel in this action. Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 29 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -25- rights … that the plaintiff seeks to vindicate’ against ‘the precise interests put forward by the State as justifications for the burden imposed by its rule,’ taking into consideration ‘the extent to which those interests make it necessary to burden the plaintiff’s rights.’” Burdick, 504 U.S. at 434 (quoting Anderson, 460 U.S. at 789). This approach applies a “flexible” sliding scale, in which “the rigorousness of [the court’s] inquiry into the propriety of a state election law depends upon the extent to which [the challenged law] burdens [voting rights].” Id. Courts calibrate this standard in each case to “[t]he precise character of the state’s action and the nature of the burden on voters.” Ne. Ohio Coal. for Homeless v. Husted (“NEOCH”), 696 F.3d 580, 591-93 (6th Cir. 2012) (quotation marks omitted). In determining how severely a measure burdens the right to vote, the pertinent question is not the extent to which it burdens those individuals impacted by it. See Crawford v. Marion Cty. Election Bd., 553 U.S. 181, 186, 191, 198, 201 (2008) (controlling op.). Anderson/Burdick does not look at the impact of the challenged provision in isolation, but within the context of the whole election scheme. Burdick, 504 U.S. at 438- 439. Individual provisions that may not be burdensome standing alone can create unconstitutional burdens when considered in light of other challenged provisions or the broader electoral context. See Republican Party of Ark. v. Faulkner Cty., 49 F.3d 1289, 1291 (8th Cir. 1995); Wood v. Meadows, 207 F.3d 708, 713 (4th Cir. 2000). In weighing competing interests, courts applying Anderson/Burdick must look skeptically upon vague or speculative state interests. A state must identify its “precise interests” in a challenged provision. Burdick, 504 U.S. at 434. a. Allocation of polling locations. The burdens imposed by the County’s allocation decisions in the 2016 PPE were many and they were severe, resulting not only in long wait times and loss of voter confidence, but also in the disenfranchisement of likely thousands of Arizona voters. See supra 2-5. These severe burdens were the direct result of the Recorder’s arbitrary and unsupportable allocation decisions which, if repeated in the General Election, run serious Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 30 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -26- risk of again severely burdening the fundamental voting rights of Arizona’s citizens. See supra at 13-16. While the County undoubtedly enjoys great discretion in making elections allocation decisions, it has no legitimate interest in disregarding plainly relevant factors to make those decisions with the result that the voting rights of its citizens, are severely burdened in some cases to the point of total disenfranchisement. As discussed, the County has not provided the public with any reason to believe that the same faulty assumptions will not again form the basis of its allocation decisions. See supra 22-23. To the contrary, it appears that, in planning for the General Election, the Recorder is again failing to consider the predictable differences in turnout within and between precincts, given that she has thus far stated that she only plans to allocate one polling location per precinct, no matter the precinct’s size or relevant statistics; it appears that she still does not understand the importance of properly estimating voter check-in time to appropriately plan to avoid lines, see Ex. 35 ; and that she is not considering projected turnout, see infra 23. Accordingly, the justifications provided by the County are plainly insufficient to overcome the severe burdens that allowing it to allocate polling locations without oversight would cause. b. Out-of-precinct provisional ballots. Since 2012 alone, Arizona has categorically disenfranchised over 14,500 voters for casting a ballot in a precinct OOP. Rodden Rpt. 26; see also Ex. 31 at 5 ; Ex. 29 at 9 . No other state comes close to disenfranchising this magnitude of voters for this reason. Id. The disparity between Arizona and other states is strong evidence that Arizona’s election administration process is largely responsible for this phenomena. Id. at 22-29. “It is simply not plausible that 11,000 Arizonans stood in line to vote in 2012 and went to the trouble of casting a provisional ballot that they know would not be counted.” Id. at 27; see also Vasquez Decl. ¶ 6; Kelso Decl. ¶ 7; Noonan Decl. ¶ 7; Quezada Decl. ¶ 25; R. Gallego Decl. ¶ 17. Individual-level analysis reveals that voter confusion caused by the large number of changes in polling locations from election to election is one of the primary factors causing voters to cast OOP ballots. See Rodden Rpt. 2, 5, 8-9, 12-14, 44, Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 31 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -27- 56-59, 62-65. “The rate of out-of-precinct voting was 40 percent higher for voters that had experienced a change in polling place.” Id. at 57. Polling location placement, inconsistent election regimes used by and within counties, and procedural errors are also contributing causes. See supra 7-9. Given the magnitude of the voters disenfranchised, as well as the strong evidence that it is Arizona’s own highly confusing elections administration that is causing its voters to cast ballots OOP at high rates, there is no legitimate justification to support this practice. Defendants’ policy of rejecting OOP ballots in jurisdictions that opt to run an election under a precinct-based system, while other jurisdictions holding the same election under a vote center based system count ballots voted anywhere in the county, further violates the 14th Amendment’s Equal Protection Clause because it treats similarly situated voters differently without sufficient justification for doing so. “[T]he right to vote is ‘the protected right, implicit in our constitutional system, to participate in state elections on an equal basis with other qualified voters.’” Ariz. Minority Coal. for Fair Redistricting v. Ariz. Indep. Redistricting Comm’n, 211 Ariz. 337, 346 (Ct. App. 2005) (quoting San Antonio Sch. Dist. v. Rodriguez, 411 U.S. 1, 35 n.78 (1973)). “Having once granted the right to vote on equal terms, the State may not, by later arbitrary and disparate treatment, value one person’s vote over that of another.” Bush v. Gore, 531 U.S. 98, 104–05 (2000). Where voters in “different counties have significantly different probabilities of having their votes counted, solely because of the nature of the system” the “system does not afford the ‘equal dignity owed to each voter.’” Black v. McGuffage, 209 F. Supp. 2d 889, 899 (N.D. Ill. 2002) (quoting Bush, 531 U.S. at 104-05). Arizona’s treatment of OOP ballots arbitrarily creates two classes of similarly situated voters in violation of the Equal Protection Clause: (1) fully eligible Arizona voters who are disenfranchised because they vote outside of their precinct in a precinct- system county, and (2) fully eligible Arizona voters whose ballots are counted no matter where they vote in a vote-center county. Thus, voters in a county such as Yuma or Yavapai, which ordinarily use a vote center model, have a significantly better probability Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 32 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -28- of having their vote counted than voters in counties such as Maricopa and Pima, which ordinarily use precinct-based systems. This is particularly arbitrary considering that in many instances candidates actually represent people across county lines. Fernandez Decl. ¶ 4. As the Ninth Circuit has explained “[t]he fact that two groups live on opposite sides of a political boundary does not necessarily mean they can be treated differently for voting purposes. . . . any “[] restriction that disenfranchises citizens based on where they live in relation to that arbitrary boundary cannot stand.” Pub. Integrity All., Inc. v. Tucson, 805 F.3d 876, 882-83 (9th Cir. 2015), reh’g en banc granted, 2016 WL 1696226 (9th Cir. Apr. 27, 2016). Arizona’s OOP practice does precisely that and, accordingly, violates the Equal Protection Clause and must be enjoined. B. C. Plaintiffs Will Suffer Irreparable Harm Absent An Injunction “It is well established that the deprivation of constitutional rights ‘unquestionably constitutes irreparable injury.’” Melendres I, 695 F.3d at 1002 (quoting Elrod v. Burns, 427 U.S. 347, 373 (1976). Courts have long recognized that restrictions on the right to vote fall into that category. LOWV, 769 F.3d at 247; Obama for Am. v. Husted, 697 F.3d 423, 436 (6th Cir. 2012) (OFA); Council of Alt. Political Parties v. Hooks, 121 F.3d 876 (3d Cir. 1997); Williams v. Salerno, 792 F.2d 323, 326 (2d Cir. 1986). These cases all recognize the indisputable fact that, “once the election occurs, there can be no do-over and no redress.” LOWV, 769 F.3d at 247. The instant case is no different. As discussed, the County’s arbitrary approach to allocation decisions and the State’s OOP policy are likely to cause Plaintiffs, their members and constituents, and thousands of other Arizona citizens irreparable harm, by imposing severe burdens on, and in some cases entirely denying, their fundamental right to vote. This factor accordingly supports the injunction. D. The Balance Of The Equities Tips Sharply In Plaintiffs’ Favor The balance of the equities also tips strongly in favor of issuing the requested injunction. See Ariz. Dream Act Coal. v. Brewer, 818 F.3d 901, 920 (9th Cir. 2016) (“The public interest and the balance of the equities favor ‘prevent[ing] the violation of a party’s Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 33 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -29- constitutional rights.’”) (quoting Melandres I, 695 F.3d. at 1002). While Defendants may incur some administrative or financial costs as a result, any such costs are far outweighed by the constitutional injury that Plaintiffs, their members and constituents, and thousands of other Arizona citizens are likely to suffer absent an injunction. See, e.g., Taylor v. La., 419 U.S. 522, 535 (1975) (holding “administrative convenience” cannot justify practices that impinge upon fundamental rights); Johnson v. Halifax Cty., 594 F. Supp 161, 171 (E.D.N.C. 1984) (“administrative and financial burdens on the defendant … are not … undue in view of the otherwise irreparable harm to be incurred by plaintiffs”). With respect to the relief requested in relation to the County’s polling allocation plan in particular, at most the County is likely to incur de minimus additional costs, given that it already has an obligation to create a polling allocation plan. See Beerheide v. Zavaras, 997 F. Supp. 1405, 1410 (D. Colo. 1998) (finding equities balanced in favor of plaintiff where plaintiffs’ First Amendment rights were threatened and requested relief would have a de minimis impact on defendants’ budget). Moreover, issuing the requested injunction is likely to confer on the County a substantial benefit in the form of the public’s increased confidence in the election system. Cf. M.R. v. Dreyfus, 697 F.3d 706, 738 (9th Cir. 2011) (balance of hardship favors plaintiffs who challenge cuts to state programs when the record suggests that the cut to be enjoined would actually cost the state money). With regard to enjoining Defendants’ practice of rejecting OOP, voters injured by this practice are completely disenfranchised, despite the fact that, in many cases, their mistake is the inevitable result of ever-changing polling locations, inconsistent rules about counting ballots across jurisdictions, inconsistent means of holding elections, or error by poll workers and elections officials. See supra at 7-9. On the other side of the scale is a policy that is not required or authorized by state law, see supra at 22-23, and the injunction of which will impose, at most, minimal administrative burdens. In particular, Plaintiffs’ seek narrow relief, requesting only that the State be required to count the races on an OOP provisional ballot for which the voter would have been eligible to vote had she cast a ballot in her assigned precinct. Plaintiffs are not requesting that the Court require all Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 34 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -30- counties use a vote center system, nor are they requesting that each precinct be supplied with every possible permutation of the ballots used in that jurisdiction. Thus, if the injunction issues, the only difference from an elections administration perspective would be that, instead of rejecting these ballots wholesale, the votes cast in races in which the voter is eligible to vote would be counted. This is not a substantial burden as counties already have processes in place to sort and count provisional ballots that are not cast OOP. Ex. 7 at 182-87. Further, because Arizona has a dual registration system in which some registered voters are only able to vote for federal candidates, Arizona already has a process in place to count federal portions of a provisional ballot when a voter is not eligible to vote for state races on the ballot. Id. at 182-83. Accordingly, where the Plaintiffs and thousands of Arizona residents face the prospect of denial and abridgment of a fundamental right, and the burdens faced by Defendants are minimal, in all instances the equities clearly balance in Plaintiffs’ favor. E. The Public Interest Favors The Issuance Of A Preliminary Injunction “‘[I]t is always in the public interest to prevent the violation of a party’s constitutional rights.’” Melendres I, 695 F.3d at 1002. It is equally well established that “[t]he public has a ‘strong interest in exercising the fundamental political right to vote.’” LOWV, 769 F.3d at 248 (quoting Purcell v. Gonzalez, 549 U.S. 1, 4 (2006)); OFA, 697 F.3d at 437 (“The public interest . . . favors permitting as many qualified voters to vote as possible.’”). If not enjoined, both the County’s irrational allocation practices and the State’s practice of not counting OOP ballots are highly likely to directly interfere with the fundamental right to vote of thousands of Arizona’s citizens, including Plaintiffs and their core constituencies, in the process decreasing the amount of voters participating in the upcoming General Election. Rodden Rpt. 14-18, 69. Accordingly, the public interest strongly weighs in favor of issuing a preliminary injunction. III. CONCLUSION For these reasons, Plaintiffs respectfully request that the Court issue the requested preliminary injunction. A proposed order is submitted herewith. Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 35 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -31- Dated: June 10, 2016 /s/ Daniel C. Barr Daniel C. Barr (# 010149) Sarah R. Gonski (# 032567) PERKINS COIE LLP 2901 North Central Avenue, Suite 2000 Phoenix, Arizona 85012-2788 Marc E. Elias (WDC# 442007)* Bruce V. Spiva (WDC# 443754)* Elisabeth C. Frost (WDC# 1007632)* Amanda R. Callais (WDC# 1021944)* PERKINS COIE LLP 700 Thirteenth Street N.W., Suite 600 Washington, D.C. 20005-3960 Telephone: (202) 654-6200 Facsimile: (202) 654-6211 MElias@perkinscoie.com BSpiva@perkinscoie.com EFrost@perkinscoie.com ACallais@perkinscoie.com Attorneys for Plaintiffs /s/ Roopali Desai Roopali H. Desai (# 024295) Andrew S. Gordon (# 003660) D. Andrew Gaona (# 028414) COPPERSMITH BROCKELMAN PLC 2800 N. Central Avenue, Suite 1200 Phoenix, Arizona 85004 Malcolm Seymour* GARVEY SCHUBERT BAKER 100 Wall Street, 20th Floor New York, New York 10005-3708 Telephone: (212) 965-4533 MSeymour@gsblaw.com Attorneys for Intervenor-Plaintiff Bernie 2016, Inc. *Admitted pro hac vice Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 36 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -32- CERTIFICATE OF SERVICE I hereby certify that on June 10, 2016, I electronically transmitted the attached document to the Clerk’s Office using the CM/ECF System for filing and a Notice of Electronic Filing was transmitted to counsel of record. /s/ Daniel R. Graziano Case 2:16-cv-01065-DLR Document 73 Filed 06/10/16 Page 37 of 37